formin methods
FORMULATING THE METHODS
Outline
How to present methods
Why planning is important
Two principles for planning experiments
Describing participants
Selecting and describing instruments
Describing procedures
Describing design and analysis
Establishing cause and effect
Interaction of participants, measurements, and treatments
HOW TO PRESENT METHODS
The purpose of the methods section is to explain how the study was conducted.
The description must be thorough enough that a competent researcher could replicate the study.
A typical methods section is divided into four parts:
1) Participants
2) Instruments or apparatuses
3) Procedures
4) Design and analysis
WHY PLANNING IS IMPORTANT
Careful planning is crucial to eliminate any alternative or rival hypotheses.
Correct study design leads to predictable results, implying the only explanation is the research conducted.
Example: Testing a hypothesis where shoe size is positively related to mathematics performance in elementary children (grades 1-5).
Students' shoe sizes are measured alongside standardized math scores.
The plotted data shows each dot representing a single student, indicating a trend: larger shoe size correlates with better math performance.
Critical reflection: "Could this be true? What have we overlooked?"
MAXICON PRINCIPLE
Remember the MAXICON principle which consists of:
Maximize true variance: Increase the odds of discovering the real relationship or explanation.
Minimize error variance: Reduce any mistakes that may obscure the true relationship.
Control extraneous variables: Ensure that rival hypotheses do not serve as real explanations for observed relationships.
TWO PRINCIPLES FOR PLANNING EXPERIMENTS
Less is more:
Refers not to the number of participants but the number of independent and dependent variables.
Avoid adding extra variables solely for exploratory purposes.
Simple is better:
Simplicity in treatment design, analysis, and displaying data is essential.
Overly complex studies can impede interpretation of results, even for the researchers.
DESCRIBING PARTICIPANTS
This section outlines how participants were selected and the pertinent characteristics for the study:
Are participants with special characteristics necessary?
Age: Specify categories such as children, adolescents, young-middle-aged adults, older adults (provide age in years).
Sex: Male, female, or both.
Level of training: Trained vs. untrained.
Level of performance: Experts vs. novices.
Size: Weight, adiposity, etc.
Special types: Defining attributes such as athletes, cyclists, sedentary individuals.
Considerations:
Can necessary permission and cooperation from participants be obtained?
Is it feasible to find enough participants?
REPORTING PARTICIPANT CHARACTERISTICS
Clearly include the exact number of participants, often noted as N = .
List participant characteristics crucial for the research, defined as needed.
For reporting, express characteristics statistically (M ± SD).
SELECTING AND DESCRIBING INSTRUMENTS
Selecting instruments, apparatuses, or tests for data collection requires careful planning.
Critical questions to consider include:
What are the validity and reliability measures? (Validity: Does it measure accurately? Reliability: Does it provide consistent measurements?)
How difficult is it to obtain the measures?
Do you have access (budget considerations) to the instruments/tests needed?
Are you capable of administering the tests or using the equipment?
Can you evaluate test performance adequately?
Will the tests yield a reasonable range of scores for selected participants?
Are participants willing to commit the necessary time for test administration?
EXAMPLE OF INSTRUMENT SELECTION
In a sport psychology study on how college athletes respond to lectures on steroid use:
Administer three tests:
Steroid knowledge test
Attitude about responsible drug use
Trait personality measures.
Describe reliability and validity of each test with citations and explain scoring methods.
DESCRIBING PROCEDURES
Detailed description of data acquisition (testing procedures and data analysis).
Organization should typically be chronological.
Include details such as:
Who administered the tests?
The testing situation and participant preparation.
Specific instructions provided to participants.
POINTS TO CONSIDER IN PROCEDURES
When collecting data, clarify:
When, where, and duration of tests.
Use of pilot data to demonstrate researcher skill with tests/equipment and understanding of participant responses.
A well-designed scheme for data acquisition, recording, and analysis is imperative.
TREATMENT PLANNING
Determine treatment specifics:
Duration, intensity, and frequency of interventions.
Methods for assessing participant adherence to treatments.
Use pilot data to preemptively gauge participant responses and researcher administration capabilities.
Ensure treatments are appropriate for the participant type.
REPLICATION IN WRITING
Procedures must detail an order of steps taken:
Timing of study, duration of procedures, and gaps between procedures.
Instructions delivered to participants requiring replication.
Include briefings, debriefings, and necessary safeguards.
DESCRIBING DESIGN AND ANALYSIS
Designing research is crucial for outcome control.
Independent variables must be manipulated to examine their impact on the dependent variable.
A well-designed experiment ensures that changes in the dependent variable solely arise from the independent variable treatment.
STATISTICAL ANALYSES
Statistical techniques should be explained:
Descriptive statistics such as means ± SD for each variable.
For correlational techniques, variables must be named alongside specific analysis methods used.
In experiments, statistics detailing between-group differences must also be presented.
ESTABLISHING CAUSE AND EFFECT
Establishing cause-and-effect relationships involves more than just design and statistics.
Statistical tests aim to either retain or reject the null hypothesis:
If rejected, what remains is the research hypothesis, though proving this can be challenging.
Criteria for establishing cause and effect:
Method of agreement: An effect occurring when both A and B exist, with C in common likely indicates C as the cause.
Method of disagreement: An effect absent in E and F when C is the only common absent element suggests C is the cause.
MANIPULATION EFFECTS IN EXPERIMENTATION
In experimental research, treatment involvement occurs regularly.
Example: Participants in a home exercise program are expected to exercise daily for a minimum of 40 minutes.
Assurance of participant adherence may require manipulation checks.
MANIPULATION CHECKS
Example manipulation check: requiring participants to wear activity monitors as validation of exercise engagement.
Identify challenges: Fraudulent reporting may occur if participants ask non-participants to wear devices.
Laboratory settings facilitate easier checks compared to real-world settings.
FATAL FLAWS OF RESEARCH
Evaluating studies for fatal flaws is essential.
Consider the following questions:
Does the study lack characteristics that lead to automatic rejection for publication regardless of methodology or outcomes?
Are hypotheses logically consistent with the theory and study characteristics?
Are assumptions about the study rational?
Is there an adequate number of "right" participants selected?
Are treatments sufficiently intense and lengthy to effect desired changes?
ADDITIONAL EVALUATION QUESTIONS
Are extraneous variables being adequately controlled?
Are dependent variables effectively characterizing participant responses to treatments?
Are all measurements both valid and reliable for the context?
Are data collection and storage procedures meticulously planned and executed?
INTERACTION OF PARTICIPANTS, MEASUREMENTS AND TREATMENTS
In both correlational and experimental studies, participant selection, measurement, and treatment choice is vital.
For correlational studies, measurements should capture the critical characteristics and yield an appropriate range of scores to facilitate discovering important relationships.
Example: To study the relationship between anxiety and motor performance, participants need a range of anxiety levels and motor skills to draw valuable conclusions.
In experimental studies, this principle also applies when evaluating the impact of a treatment (independent variable) on dependent variables.
Example: Determining the influence of a resistance training program on jumping performance requires diverse baselines in participants’ training backgrounds and skills to enable observation of treatment effects.